[WCCFTech]AMD’s FirePro Professional GPUs compared to Quadro counterparts

[Dribble]

Im sure you can provide a proof for those claims ??

I work for a company that produces high end engineering software (million+ users), I sit about 20 foot from the guy writing the renderer, that is my personal experience. Nvidia > AMD > Intel.

 

[RussianSensation]

I work for a company that produces high end engineering software (million+ users), I sit about 20 foot from the guy writing the renderer, that is my personal experience. Nvidia > AMD > Intel.

A friend of mine just went in to BMW service dealer with his X6M with only 80,050 kms:

8 fuel injectors broken
8 fuel pumps broken
Variable Valve broken
High pressure sensors broken
Cost to repair $9118.

Would it be reasonable to conclude from this anecdotal evidence that all BMWs fall apart like this and that all BMW M5/M6/X5M/X6M/6 Series M GranCoupe/6 Series M Coupe/Convertible that have the same engine will all fall apart like this at 80000 kms too?

Having said that, there are some instances of driver bugs/glitches with Firepros (just experienced one a few weeks ago). Employees might suffer downtime due to getting these sorted out. Quadros typically don't have any issues and the lack of downtime is less. Downtimes cost alot of money so at the end of the day, the prices could actually even out.

Or this scenario:

I need an all-purpose card for scientific compute calculations and workstation usage:

W9100 = $3100 - 50% off = $1550.

With that I'd get 5.2Tflops of SP performance and 2.62Tflops of DP performance.

vs. Nvidia option:

Workstation = K5200 for $1900

Compute Option 1 = K40 = 4.29Tflops of SP performance and 1.43Tflops of DP performance = $3300

Compute Option 2 = K80 = 8.74Tflops of SP performance and 2.91 Tflops of DP performance = $5500

So my options are paying $1550 for AMD vs. $5200-7400 for Option 1 or 2 NV. Let's say I need 10-20 of those cards. Maybe a large corporation doesn't care between spending $15,000 and $52,000 but I am sure smaller businesses do care. For universities performing scientific computing with DP, the $1550 W9100 is a bargain against the $5500 K80. Think about how many more FirePros you could purchase to speed up your calculations.

 

[Cookie Monster]

And say you have 3 days downtime because of some stupid driver glitch. Thats money wasted on employees twiddling their thumb or trying to troubleshoot their drivers. What about the wasted project cost?

Im not even saying nVIDIA is perfect either. But when it comes to professional workstation cards, they have a reputation for a reason. Just like rhodes and schwarz (they cost $$$$), agilent technologies, weller-tools etc compared to its lower price competitors.

Large corporations do care and is the reason why they spend more so that it doesn't come back to haunt them or negatively affect critical projects. Reason why many electronic companies also invest in alot of money on ESD protection systems/tools even if there is no scientific data to back up exactly how prone electronics are due to ESDs (and at what rate). Similar to an earthquake insurance. Yet it can result in millions of dollars worth of damages IF it happens.

 

[Enigmoid]

I need an all-purpose card for scientific compute calculations and workstation usage:

W9100 = $3100 - 50% off = $1550.

With that I'd get 5.2Tflops of SP performance and 2.62Tflops of DP performance.

vs. Nvidia option:

Workstation = K5200 for $1900

Compute Option 1 = K40 = 4.29Tflops of SP performance and 1.43Tflops of DP performance = $3300

Compute Option 2 = K80 = 8.74Tflops of SP performance and 2.91 Tflops of DP performance = $5500

So my options are paying $1550 for AMD vs. $5200-7400 for Option 1 or 2 NV. Let's say I need 10-20 of those cards. Maybe a large corporation doesn't care between spending $15,000 and $52,000 but I am sure smaller businesses do care. For universities performing scientific computing with DP, the $1550 W9100 is a bargain against the $5500 K80. Think about how many more FirePros you could purchase to speed up your calculations.

RS's logic is sound as always. But based on an incorrect premise and thus not valid.

Its 50% off for the FIRST W9100. If you buy 10-20 of them you pay full price for 9-19 and half price for the first.

AMD reviews your application for rebate and may or may not give you the rebate.

In the context of buying 10-20 of them the price difference is pretty minimal.

Model
Idle
3D Workload, Open Lab Bench
3D Workload, Closed Case
Quadro K5000
30.8 dB(A)
37.7 dB(A)
37.1 dB(A)
Quadro K6000
30.8 dB(A)
42.7 dB(A)
41.2 dB(A)
FirePro W9100
33.5 dB(A)
51.3 dB(A)
49.8 dB(A)
FirePro W9000
33.2 dB(A)
55.4 dB(A)
52.7 dB(A)

With the reference cooler this thing if very loud. Not as bad a the W9000 but nearly 9 dB louder than the K6000. As a professional that is not something you want on your desk. For universities that is not something you want 30x in the lab room.

 

[Silverforce11]

With the reference cooler this thing if very loud. Not as bad a the W9000 but nearly 9 dB louder than the K6000. As a professional that is not something you want on your desk. For universities that is not something you want 30x in the lab room.

When Apple announced Macs with FirePros the first thing I thought of, who the heck wants to work next to the reference Hawaii blower?!

I couldn't deal with it tbh, which was why I went with water cooling. But apparently lots of people don't mind, hence the significant marketshare gains!

 

[Dribble]

A friend of mine just went in to BMW service dealer with his X6M with only 80,050 kms:

8 fuel injectors broken
8 fuel pumps broken
Variable Valve broken
High pressure sensors broken
Cost to repair $9118.

Would it be reasonable to conclude from this anecdotal evidence that all BMWs fall apart like this and that all BMW M5/M6/X5M/X6M/6 Series M GranCoupe/6 Series M Coupe/Convertible that have the same engine will all fall apart like this at 80000 kms too?
.

I do not work for a shop that sells and repairs graphics cards, we write software which requires working drivers from Nvidia/AMD/Intel, and support from Nvidia/AMD/Intel.

 

[96Firebird]

Are you seriously asking us to link you 23 data points (Q1 2009 to Q3 2014)?

Yes, that is exactly what I am asking. Something like this...

But for professional cards only.

 

[Enigmoid]

When Apple announced Macs with FirePros the first thing I thought of, who the heck wants to work next to the reference Hawaii blower?!

I couldn't deal with it tbh, which was why I went with water cooling. But apparently lots of people don't mind, hence the significant marketshare gains!

Apple is not using reference designs.

 

[Silverforce11]

Apple is not using reference designs.

IIRC the new FirePros have two cooler variants, a blower thats identical to consumer R290X and a large copper heat chamber design without a fan, reliant on the HPC cluster airflow.

You got a link to the Apple design?

 

[Gloomy]

IIRC the new FirePros have two cooler variants, a blower thats identical to consumer R290X and a large copper heat chamber design without a fan, reliant on the HPC cluster airflow.

You got a link to the Apple design?

It's a large triangular heatsink. https://www.ifixit.com/Teardown/Mac+Pro+Late+2013+Teardown/20778

 

[Silverforce11]

It's a large triangular heatsink. https://www.ifixit.com/Teardown/Mac+Pro+Late+2013+Teardown/20778

That small heatsink and one fan is enough for two Hawaii and a CPU too? Wow. Apple magic.

Edit: Ok, those are entry models with lower end GPUs. For a moment, I thought they went full 2x Hawaii on a small ITX compact rig with a 450W PSU... heh

 

[RussianSensation]

That small heatsink and one fan is enough for two Hawaii and a CPU too? Wow. Apple magic.

Edit: Ok, those are entry models with lower end GPUs. For a moment, I thought they went full 2x Hawaii on a small ITX compact rig with a 450W PSU... heh

It's up to a maximum of dual Tahiti XT, clocked at 850mhz though:
http://www.anandtech.com/show/7603/mac-pro-review-late-2013/9

Apple announced the Mac Pro on June 10, 2013, which would have made it impossible to have R9 series on it:
http://www.engadget.com/2013/06/10/apple-mac-pro-2013-redesign/

For the next upgrade they will probably go with 300 series or Maxwell. Hawaii is already 1 years old. If Apple updates the Mac Pro in 2015, they will want the latest tech. Perhaps announce the update June 2015, with product availability in the Fall 2015.

 

[NTMBK]

That small heatsink and one fan is enough for two Hawaii and a CPU too? Wow. Apple magic.

Edit: Ok, those are entry models with lower end GPUs. For a moment, I thought they went full 2x Hawaii on a small ITX compact rig with a 450W PSU... heh

Small heatsink? It's freaking huge! And the design of the chassis means that it is always cool external air which enters, and hot air is immediately vented instead of percolating around and heating up other components.

 

[RussianSensation]

Yes, that is exactly what I am asking. Something like this...

But for professional cards only.

As I said already it's very difficult to find per quarter market share for professional GPUs unless you have subscription to the top equity research reports covering the industry or spend time looking at every quarterly report and try to compile the data yourself (if it's there).

The trend that AMD is slowly gaining professional GPU market share is there.

AMD has also worked hard to reduce the number of reported driver problems per 1 million graphics cards sold, even though the perception that AMD drivers are bad still remains.

Some analysts project that AMD could hit 30% by end of 2014. Even if AMD keeps gaining 3-4% a year, in 5 years they'll be close to 50%.

 

[Vesku]

u are wasting time.They Dont judge performance and e.t.c on facts because they have different method with no source and links.

Or perhaps certain facts change such as price?

From the xbit labs article: "FirePro W8100 (priced at about $2300)"

Current Newegg W8100 price $1099.99.

AMD has more to prove in this market than the gaming GPU one, Nvidia is quite entrenched especially in terms of software integration. More so than the gaming GPU market. Business customers will want to hear good things and have good experiences when it comes to support and bug fixes. Although depending on the business that can be as basic as partnering with the software maker to get some prominent "Designed for FirePro" badging.

 

[96Firebird]

LOL at using SA as a source. I appreciate the effort, but I've seen too many completely made up articles to believe anything from that site.

 

[wand3r3r]

LOL at using SA as a source. I appreciate the effort, but I've seen too many completely made up articles to believe anything from that site.

RS has provided a couple sources, more than enough to support his statement. Feel free to google the quarterly data if you think it disproves the data presented but why should someone else do the research for you?

 

[RussianSensation]

LOL at using SA as a source. I appreciate the effort, but I've seen too many completely made up articles to believe anything from that site.

There have been 2-3 sources linked in this thread already that all corroborate the same data that AMD's market share in professional graphics went from 12% in 2009 to near 25% by Q3 2014. It's up to you to believe that or not. It doesn't guarantee that AMD will continue growing their professional market share at the same pace, if at all, but it shows a major change in the market because for 10+ years ATI couldn't even make a dent.

Ideally we would want a 50-50 market share so that competition is more fierce but it's a good start already from where AMD graphics were in 2009.

 

[96Firebird]

I'm not looking for that data, frankly I don't care about AMD's current marketshare in professional graphics. If you look back at my posts, you'll see why I want the data.

Don't worry guys, this isn't always about AMD vs Nvidia! I know you all can't help yourselves to make it that way though.

 

[RussianSensation]

If you don't care to see market share data, then why do you keep asking for a quarterly breakdown of market share for the last 5 years? The point is you came off as questioning whatever data is provided that paints AMD in a positive light. No one stops you from looking for this information yourself. It's not like we have $5000 research reports handy.

 

[Keysplayr]

If you don't care to see market share data, then why do you keep asking for a quarterly breakdown of market share for the last 5 years? The point is you came off as questioning whatever data is provided that paints AMD in a positive light. No one stops you from looking for this information yourself. It's not like we have $5000 research reports handy.

But you kind of post like you do. Just saying. Maybe that's why he thinks you have immediate access to these reports because of that reason. You sound so sure of what you're saying, that maybe he believes you have the deets ready at hand right in front of you.

 

[3DVagabond]

But you kind of post like you do. Just saying. Maybe that's why he thinks you have immediate access to these reports because of that reason. You sound so sure of what you're saying, that maybe he believes you have the deets ready at hand right in front of you.

Or maybe he's just trying to cast doubt on the data given rather than try and provide data to the contrary because, you can't provide that which doesn't exist.

 

[DooKey]

Why cast doubt on data if contrary data doesn't exist?

 

[monstercameron]

Why cast doubt on data if contrary data doesn't exist?

to fit a certain narrative.

 

[96Firebird]

If you don't care to see market share data, then why do you keep asking for a quarterly breakdown of market share for the last 5 years? The point is you came off as questioning whatever data is provided that paints AMD in a positive light. No one stops you from looking for this information yourself. It's not like we have $5000 research reports handy.

Thanks, but I am more interested in a long-term chart that shows marketshare for each quarter. Want to compare changes with new product releases and see if there is any correlation.

Did you even read the thread?

When you said not anyone has that information to publish, I was going to leave it at that. That is understandable. But then you went on to try to prove (to yourself, presumably) that AMD is gaining share. Like I said, that doesn't concern me.